Measure Package Similarity

Measure Package Similarity

Usage

get_similar_packages(pkg, pkg_list)

Arguments

pkg

A single package name.

pkg_list

Value of a call to get_all_biocpkglist().

Value

A tibble of two columns: package and similarity. package is the name of every other package. similarity is a measure of similarity (see Details).

Details

Currently, similarity is quantified by taking the Hamming distance over the set of biocViews used to tag either of the two packages being tested when computing each distance, dividing by the number of biocViews to normalise within the range [0, 1], and finally taking one minus that value to yield a measure of similarity in the range [0, 1]. In other words, the denominator in different for each pair of packages, to increase the dynamic range of values.

Examples

biocpkglist <- get_all_biocpkglist(verbose = FALSE)

get_similar_packages(
  pkg = "edgeR",
  pkg_list = biocpkglist
)
#> # A tibble: 3,752 × 2
#>    package           similarity
#>    <chr>                  <dbl>
#>  1 metaseqR2              0.674
#>  2 limma                  0.627
#>  3 roastgsa               0.615
#>  4 dearseq                0.537
#>  5 dreamlet               0.512
#>  6 crumblr                0.488
#>  7 variancePartition      0.488
#>  8 zenith                 0.488
#>  9 GenomicOZone           0.478
#> 10 BPRMeth                0.476
#> # ℹ 3,742 more rows